The COVID-19 pandemic has transformed learning environments, challenging educators to understand students' behaviors during the online mode of learning, in particular emotions associated with students' attention during virtual classrooms. As learning transitions between physical and virtual spaces, the ability to interpret student attention and engagement has become complex. In response to this challenge, our research investigates the use of GPT-4o, a multimodal large language model, for identifying student emotions by analyzing images in diverse learning settings. The study involved analyzing online classroom images featuring 149 faces, utilizing three distinct approaches: a computer vision model (YOLO), the multimodal LLM (GPT-4o), and a human-annotated baseline. The analysis systematically categorized facial expressions into eight emotional categories: Happy, Sad, Angry, Neutral, Contempt, Disgust, Fear, and Surprise. The findings indicate that multimodal LLMs can effectively detect student emotions, achieving an average accuracy of 93.8%, which aligns with the human baseline accuracy of 97.0%. In contrast, YOLO models maintained an average accuracy of 81.9%, performing well for basic emotions but struggling with subtle expressions. This research contributes to enhancing educational practices by providing valuable insights regarding the application of multimodal LLMs to assist educators in comprehending student emotions within both physical and digital classroom settings.

GitHub Actions (GHA) has become among the most popular Continuous Integration (CI) platforms in open-source software (OSS) and commercial projects. Collecting such build data remains crucial for practitioners and researchers to allow build performance monitoring, optimization and improvement. However, mining GHA builds to collect build-related data and metrics remains challenging and time-consuming. This paper introduces GHAminer, an open-source tool designed to collect build-related metrics for GitHub Actions. GHAminer covers various aspects of data such as the build-related code changes and tests, the build duration and status (e.g., passed, failed, timeout, etc.), and repository metadata, which would be useful for practitioners and researchers to make data-driven decisions to enhance CI efficiency and quality. The tool has a modular architecture that supports efficient data extraction with minimal API load. Specifically, it consists of a set of modules that are related to repository information collection, build analysis, commit history analysis, and build log parsing. We evaluate the performance of GHAminer on a representative sample of 3,151 OSS projects. Results show that GHAminer is efficient in handling projects of various sizes with relatively stable performance to collect build data for larger projects. GHAminer is publicly available with a demo video at: https:lIgithub.com/stilab-ets/GHAminer

Traffic Signal Control (TSC) is a crucial component in Intelligent Transportation Systems (ITS) for optimizing traffic flow. Deep Reinforcement Learning (DRL) techniques have emerged as leading approaches for TSC due to their promising performance. Most existing DRL-based approaches typically use discrete action spaces to predict the next action phase, without specifying the signal duration. In contrast, some studies employ continuous action spaces to determine signal phase timing within a fixed light cycle. To address the limitations of both approaches, we propose a flexible framework that predicts both the appropriate traffic light phase along with its associated duration. Our approach utilizes a Parameterized-action based deep reinforcement learning architecture to handle the combination of discrete-continuous actions. We evaluate our method using the Simulation of Urban MObility (SUMO) environment, comparing its efficiency against state-of-the-art techniques. Results demonstrate that our approach significantly outperforms traditional and learning-based methods.

Intelligent Transportation Systems are essential due to the increased number of traffic congestion problems and challenges nowadays. Traffic Signal Control (TSC) plays a critical role in optimizing the traffic flow and mitigating the congestion within the urban areas. Various research works have been conducted to enhance the behavior of TSCs at intersections and subsequently reduce the traffic congestion. Researchers recently leveraged Deep Learning (DL) and Reinforcement Learning (RL) techniques to optimize TSCs. In RL framework, the agent interacts with surrounding world through states, rewards and actions. The formulation of these key elements is crucial as they impact the way the RL agent behaves and optimizes its policy. However, most of existing frameworks rely on hand-crafted state and reward designs, restricting the RL agent from acting optimally. In this paper, we propose a novel approach to better formulate state and reward definitions in order to boost the performance of the traffic signal controller agent. The intuitive idea is to define both state and reward in a consistent and straightforward manner. We advocate that such a design approach helps achieving training stability and hence provides a rapid convergence to derive best policies. We consider the double deep Q-Network (DDQN) along with prioritized experience replay (PER) for the agent architecture. To evaluate the performance of our approach, we conduct series of simulations using the Simulation of Urban MObility (SUMO) environment. The statistical analysis of our results show that the performance of our proposal outperforms the state-of-the-art state and reward design approaches.

The rapid growth of mobile applications development and usage raises several new challenges to developers as they need to respond quickly to the users’ needs in a world of continuous changes. Developers often use third-party libraries to add functionality, which significantly improves developers productivity, and reduces time-to-market. In this paper, we present an approach for the visualization and recommendation of libraries for Android apps. Our approach, named LibScanDroid, is based on how libraries are used within existing Android applications. LibScanDroid groups together libraries based on their history of joint and separate usage in existing Android applications available in Google Play Store. The library groups, i.e., usage patterns, are presented in several layers to visualize and navigate through the patterns. These groupings are performed using the ϵ-DBSCAN hierarchical clustering algorithm.We implement our approach in the form of an interactive tool and evaluate it on a database that covers 1,458 libraries that are used by over 1,000 Android applications. Our experiments have shown that our approach can detect library patterns with high co-usage cohesion. The results from the cross-validation, allows us to affirm the generalizability of the detected patterns.

Web service composition allows developers to create applications via reusing available services that are interoperable to each other. The process of selecting relevant Web services for a composite service satisfying the developer requirements is commonly acknowledged to be hard and challenging, especially with the exponentially increasing number of available Web services on the Internet. The majority of existing approaches on Web Services Selection are merely based on the Quality of Service (QoS) as a basic criterion to guide the selection process. However, existing approaches tend to ignore the service design quality, which plays a crucial role in discovering, understanding, and reusing service functionalities. Indeed, poorly designed Web service interfaces result in service anti-patterns, which are symptoms of bad design and implementation practices. The existence of anti-pattern instances in Web service interfaces typically complicates their reuse in real-world service-based systems and may lead to several maintenance and evolution problems. To address this issue, we introduce a new approach based on the Multi-Objective and Optimization on the basis of Ratio Analysis method (MOORA) as a multi-criteria decision making (MCDM) method to select Web services based on a combination of their (1) QoS attributes and (2) QoS design. The proposed approach aims to help developers to maintain the soundness and quality of their service composite development processes. We conduct a quantitative and qualitative empirical study to evaluate our approach on a Quality of Web Service dataset. We compare our MOORA-based approach against four commonly used MCDM methods as well as a recent state-of-the-art Web service selection approach. The obtained results show that our approach outperforms state-of-the-art approaches by significantly improving the service selection quality of top- k selected services while providing the best trade-off between both service design quality and desired QoS values. Furthermore, we conducted a qualitative evaluation with developers. The obtained results provide evidence that our approach generates a good trade-off for what developers need regarding both QoS and quality of design. Our selection approach was evaluated as “relevant” from developers point of view, in improving the service selection task with an average score of 3.93, compared to an average of 2.62 for the traditional QoS-based approach.

During COVID-19 pandemic time, the educational systems has been enormously transformed. Indeed, the outbreak has led to complete shutdown of schools, colleges and universities around the globe, in order to curb the transmission of coronavirus. The new circumstances of learning have made digital learning or distance learning a solution to preserve the mission of educational system. Instructors, professors and students, have found themselves working in completely new conditions. While the global shutdown of educational institutions have caused a crucial interference in students learning, worldwide, digital innovations and emerging technologies such as artificial intelligence have captured the interest of many researchers in order to support the education system during this era of pandemic. In this paper, we investigate the role of artificial intelligence in preserving the mission of learning operation during the outbreak while applying measures and protocols to curb the pandemic. In particular, we classify the role of AI into three categories, namely descriptive, predictive and perspective analysis contributions. Among our study outcomes is a set of lessons learned from reported experience of using AI techniques and tools in learning systems during the pandemic.

A key success of a Web service is to appropriately design its interface to make it easy to consume and understand. In the context of service-oriented computing (SOC), the service’s interface is the main source of interaction with the consumers to reuse the service functionality in real-world applications. The SOC paradigm provides a collection of principles and guidelines to properly design services to provide best practice of third-party reuse. However, recent studies showed that service designers tend to pay little care to the design of their service interfaces, which often lead to several side effects known as antipatterns . One of the most common Web service interface antipatterns is to expose a large number of semantically unrelated operations, implementing different abstractions, in one single interface. Such bad design practices may have a significant impact on the service reusability, understandability, as well as the development and run-time characteristics. To address this problem, in this article, we propose a hybrid approach to improve the design quality of Web service interfaces and fix antipatterns as a combination of both deterministic and heuristic-based approaches. The first step consists of a deterministic approach using a graph partitioning-based technique to split the operations of a large service interface into more cohesive interfaces, each one representing a distinct abstraction. Then, the produced interfaces will be checked using a heuristic-based approach based on the non-dominated sorting genetic algorithm (NSGA-II) to correct potential antipatterns while reducing the interface design deviation to avoid taking the service away from its original design. To evaluate our approach, we conduct an empirical study on a benchmark of 26 real-world Web services provided by Amazon and Yahoo. Our experiments consist of a quantitative evaluation based on design quality metrics, as well as a qualitative evaluation with developers to assess its usefulness in practice. The results show that our approach significantly outperforms existing approaches and provides more meaningful results from a developer’s perspective.

Design defects are symptoms of poor design and implementation solutions adopted by developers during the development of their software systems. While the research community devoted a lot of effort to studying and devising approaches for detecting the traditional design defects in object-oriented (OO) applications, little knowledge and support is available for an emerging category of Web service interface design defects. Indeed, it has been shown that service designers and developers tend to pay little attention to their service interfaces design. Such design defects can be subjectively interpreted and hence detected in different ways. In this paper, we propose a novel approach, named WS3D, using machine learning techniques that combines Support Vector Machine (SVM) and Simulated Annealing (SA) to learn from real world examples of service design defects. WS3D has been empirically evaluated on a benchmark of Web services from 14 different application domains. We compared WS3D with the state-of-theart approaches which rely on traditional declarative techniques to detect service design defects by combining metrics and threshold values. Results show that WS3D outperforms the the compared approaches in terms of accuracy with a precision and recall scores of 91% and 94%, respectively.

In the service-oriented paradigm, Web service interfaces are considered contracts between Web service subscribers and providers. The structure of service interfaces has an extremely important role to discover, understand, and reuse Web services. However, it has been shown that service developers tend to pay little care to the design of their interfaces. A common design issue that often appears in real-world Web services is that their interfaces lack cohesion, i.e., they expose several operations that are often semantically unrelated. Such a bad design practice may significantly complicate the comprehension and reuse of the services functionalities and lead to several maintenance and evolution problems. In this paper, we propose a new approach for Web service interface decomposition using a Formal Concept Analysis (FCA) framework. The proposed FCA-based approach aims at identifying the hidden relationships among service operations in order to improve the interface modularity and usability. The relationships between operations are based on cohesion measures including semantic, sequential and communicational cohesion. The identified groups of semantically related operations having common properties are used to define new cohesive and loosely coupled service interfaces. We conducted a quantitative and qualitative empirical study to evaluate our approach on a benchmark of 26 real world Web services provided by Amazon and Yahoo. The obtained results show that our approach can significantly improve Web service interface design quality compared to state-of-the-art approaches.

The wealth of sensor data generated by advanced vehicular sensors that are fitted in new, connected vehicles enables new applications for driver behavior, road health monitoring, and incident reporting. However, standard access mechanisms to the data restricts the services and insights that can be provided by vehicular applications. For those applications to have full access to all of the vehicular sensory data, custom hardware fitted with proprietary automaker software is needed to process raw sensor values. This limits rapid deployment at scale because of the time and costs needed for mass development across proprietary platforms. In this paper, we propose a system to provide access to the raw sensory data using the vehicle's identification number in order to retrieve the vehicles sensors identification numbers, their description, and their related software libraries that house the data processing algorithms specific to the vehicle's make and model. Smartphones can collect raw sensory data through the vehicle's CAN-Bus interface, and use those software libraries to transform raw data into standard formats that can be used by vehicular applications. This way, many applications can be developed without having to worry about customization of hardware to process the data produced by each automaker's sensor platforms.

The Profile of Mood States (POMS) and its variations have been used in many real world contexts to assess individuals behavior and measure mood. Social Networks such as Twitter and Facebook are considered precious research sources of collecting user mood measurements. In particular, we are inspired in this paper, by recent work on the prediction of the stock market movement from attributes representing the public mood collected from Twitter. In this paper, we build a new prediction model for the same stock market problem based on single models combination. Our proposed approach to build such model is simultaneously promoting performance and interpretability. By interpretability, we mean the ability of a model to explain its predictions. We implement our approach using Ant Colony Optimization algorithm and we use customized Bayesian Classifiers as single models. We compare our approach against the best Bayesian single model, model learned from all the available data, bagging and boosting algorithms. Test results indicate that the proposed model for stock market prediction performs better than those derived by alternatives approaches.

Standard specifications of Web Services are mainly concerned with Web Service publishing and discovery. However, there is no standard regarding Web Service selection, which is very crucial for providers and clients. It will support clients in selecting Web Services based on required Quality of Web Service (QoWS) and support providers to remain competitive. Most of solutions on Web Service selection are very often based on a central component to make the selection decision, do not scale to the growing number of clients and Web Service providers, and/or lack trustworthiness. The objective of our approach is to support clients in selecting, and monitoring appropriate Web Services while increasing scalability, trust and reliability of Web Service selection. We propose a framework based on a federation of cooperative brokers. Each broker in the federation manages Web Services within its domain of expertise, and cooperates with its peers to select appropriate Web Services. We describe the federation management operations and our cooperative brokers QoS-aware selection algorithm. We propose certification and monitoring of Web Services as well as monitoring of brokers. The implementation and the experiments we have conducted evaluated the performance of our approach. The obtained results show that: the load is shared and distributed among brokers, a large number of client requests with different QoWS requirements were served, the architecture can scale up in order to include other brokers and/or other Web Services, and selection is guaranteed, trustworthy as it is supported by monitoring.

Abstract In a wireless network, mutual secure neighbor verification (MSN) is defined as the capability of a node (verifier) to verify the claim by another node (claimer) that it exists within a certain physical distance from the verifier. This problem has received great attention because it has numerous practical applications. Current state‐of‐the‐art approaches to solve this problem, such as the use of time of flight, signal strength, and angle of arrival, suffer from impracticality in terms of application and computation cost. In this work, we propose two algorithms to mitigate the MSN problem during the incremental deployment phase of static senor networks. Each node should announce its location and the power level it uses for transmission. Cooperative and base station verifications are used to detect nodes that lie about their locations. The simulation results show that we can achieve high detection (>90%) of nodes that forge their location information using either high power transmission or by colluding with other malicious nodes. Copyright © 2011 John Wiley & Sons, Ltd.

A transmembrane helix (TMH) topology prediction is becoming a central problem in bioinformatics because the structure of TM proteins is difficult to determine by experimental means. Therefore, methods which could predict the TMHs topologies computationally are highly desired. In this paper we introduce TMHindex, a method for detecting TMH segments solely by the amino acid sequence information. Each amino acid in a protein sequence is represented by a Compositional Index deduced from a combination of the difference in amino acid appearances in TMH and non-TMH segments in training protein sequences and the amino acid composition information. Furthermore, genetic algorithm was employed to find the optimal threshold value to separate TMH segments from non-TMH segments. The method successfully predicted 376 out of the 378 TMH segments in 70 testing protein sequences. The level of accuracy achieved using TMHindex in comparison to recent methods for predicting the topology of TM proteins is a strong argument in favor of our method.

Web Services are considered an attracting distributed approach of application/services integration over the Internet. As the number of Web Services is exponentially growing and expected to do so for the next decade, the need for categorizing and/or classifying Web Services is very crucial for their success and the success of the underlying SOA. Categorization aims at systematizing Web Services according to their functionalities and their Quality of Service attributes. Communities of Web Services have been used to connect Web Services based on their functionalities. In this paper, we augment the community approach by defining a new community to monitor Web Services operating in any Web Services community. This will be of prime importance for communities' creators/managers willing to protect and sustain their communities. This paper defines the overall architecture of the monitoring community and the basic services it offers to its various classes of customers.

Software product line engineering is an inter-disciplinary concept. It spans over the dimensions of business, architecture, process and organization. Some of the potential benefits of this approach include cost reduction, improvement in quality and a decrease in product development time. Employees' participation, organizational behavior and management contemplation play a vital role in successfully institutionalizing software product lines in a firm. Organizational dimension has been weighted as one of the critical dimensions in software product line theory and practice. Organizational behavior covers organizational culture, organizational commitment and organizational learning. A comprehensive empirical investigation to study the relationship of some organizational behavior on the performance of software product line practice is presented in this work. The results of this investigation provide empirical evidence and further support the theoretical foundations that in order to institutionalize software product lines within an organization, organizational behavior play an important role.

Fault-based testing is often advocated to overcome limitations ofother testing approaches; however it is also recognized as beingexpensive. On the other hand, evolutionary algorithms have beenproved suitable for reducing the cost of data generation in the contextof coverage based testing. In this paper, we propose a newevolutionary approach based on ant colony optimization for automatictest input data generation in the context of mutation testingto reduce the cost of such a test strategy. In our approach the antcolony optimization algorithm is enhanced by a probability densityestimation technique. We compare our proposal with otherevolutionary algorithms, e.g., Genetic Algorithm. Our preliminaryresults on JAVA testbeds show that our approach performed significantlybetter than other alternatives.

Software evolution and software quality are ever changing phenomena. As software evolves, evolution impacts software quality. On the other hand, software quality needs may drive software evolution strategies.This paper presents an approach to schedule quality improvement under constraints and priority. The general problem of scheduling quality improvement has been instantiated into the concrete problem of planning duplicated code removal in a geographical information system developed in C throughout the last 20 years. Priority and constraints arise from development team and from the adopted development process. The developer team long term goal is to get rid of duplicated code, improve software structure, decrease coupling, and improve cohesion.We present our problem formulation, the adopted approach, including a model of clone removal effort and preliminary results obtained on a real world application.

In this paper, we propose an approach for the combination and adaptation of software quality predictive models. Quality models are decomposed into sets of expertise. The approach can be seen as a search for a valuable set of expertise that when combined form a model with an optimal predictive accuracy. Since, in general, there will be several experts available and each expert will provide his expertise, the problem can be reformulated as an optimization and search problem in a large space of solutions.We present how the general problem of combining quality experts, modeled as Bayesian classifiers, can be tackled via a simulated annealing algorithm customization. The general approach was applied to build an expert predicting object-oriented software stability, a facet of software quality. Our findings demonstrate that, on available data, composed expert predictive accuracy outperforms the best available expert and it compares favorably with the expert build via a customized genetic algorithm.

Managing the software evolution for large open source software is a major challenge. Some factors that make software hard to maintain are geographically distributed development teams, frequent and rapid turnover of volunteers, absence of a formal means, and lack of documentation and explicit project planning. In this paper we propose remote and continuous analysis of open source software to monitor evolution using available resources such as CVS code repository, commitment log files and exchanged mail. Evolution monitoring relies on three principal services. The first service analyzes and monitors the increase in complexity and the decline in quality; the second supports distributed developers by sending them a feedback report after each contribution; the third allows developers to gain insight into the "big picture" of software by providing a dashboard of project evolution. Besides the description of provided services, the paper presents a prototype environment for continuous analysis of the evolution of GRASS, an open source software.

During the last ten years, a large number of quality models have been proposed in the literature. In general, the goal of these models is to predict a quality factor starting from a set of direct measures. The lack of data behind these models makes it hard to generalize, cross-validate, and reuse existing models. As a consequence, for a company, selecting an appropriate quality model is a difficult, non-trivial decision. In this paper, we propose a general approach and a particular solution to this problem. The main idea is to combine and adapt existing models (experts) in such a way that the combined model works well on the particular system or in the particular type of organization. In our particular solution, the experts are assumed to be decision tree or rule-based classifiers and the combination is done by a genetic algorithm. The result is a white-box model: for each software component, not only does the model give a prediction of the software quality factor, it also provides the expert that was used to obtain the prediction. Test results indicate that the proposed model performs significantly better than individual experts in the pool.

The goal of quality models is to predict a quality factor starting from a set of direct measures. Selecting an appropriate quality model for a particular software is a difficult, non-trivial decision. In this paper, we propose an approach to combine and/or adapt existing models (experts) in such way that the combined/adapted model works well on the particular system. Test results indicate that the models perform significantly better than individual experts in the pool.